Active transport of anionic and cationic drugs plays a critical role in their intestinal absorption, penetration across the blood-cerebrospinal fluid (CSF) and blood-brain barriers, and accumulation within excretory organs. In general, separate systems mediate the transport of charged organic solutes, the organic anion (OAT) and organic cation (OCT) transporter families. However, a number of drugs, e.g., the cation cimetidine, are transported by both systems. We have focused on the handling of the nucleoside phosphonate antiviral (NPA) drugs, using adefovir, cidofovir, and tenofovir as models. The NPAs are transported very well by OAT1 and poorly by OAT3, such that in kidney the vast majority of uptake, and thus of resulting nephrotoxicity, is mediated by OAT1. As a consequence of the limited transport of NPAs by OAT3, there is very little NPA transport into the choroid plexus or blood-brain-barrier, since these epithelia express OAT3 very well, but have little OAT1. This pattern of specificity also makes the NPAs valuable as a probes for OAT1 function in intact tissues where both OAT and OAT3 are expressed. This property of these drugs was used in study assessing the evolution of the OATs, showing that lower vertebrates have a single OAT that transports both mammalian OAT1 (NPAs)and OAT3 (estrone sulfate) substrates in kidney and barrier tissues that is apparently similar to the ancient gene that gave rise to the basolateral OATs 1 and 3 found in the mammalian epithelia. Our other current focus is on the structural features of OAT1 and OAT3 that underlie their ability to descriminate between the NPAs and other OAT substrates shared by both isoforms. Molecular modeling based on the published crystal structures of related transport proteins have provided the basis for modeling of the binding pocket in OATs 1 and 3 and identification of amino acid residues putatively involved in substrate recognition and binding. Modification of these residues by site directed mutagenisis has demonstrated their importance in transporter function and provides documentation of the predictive value of the structural model. Interestingly, although kinetic parameters for high affinity substrates like p-aminohippurate were unchanged in some of the mutants, Km values for NPAs were markedly different. Together these data indicate that the model does an effective job in highlighting key residues within the binding pocket of hOAOT1. Thus, the model (JBC, revision under review) will provide a systematic means to identify critical residues that impact the functional properties of these drug transporters.